This study examines the association between employee stock ownership plans and corporate income tax burden in the unique context of an emerging market, i.e., China, and further attempts to examine the moderating mechanism.
Using a comprehensive sample of Chinese A-share listed firms over the period 2014-2020 with 24,683 firm-year observations, this study performs panel data regression techniques as the baseline methodology. The study also employs different sensitivity tests and endogeneity tests (including propensity score matching approach) and fixed effects. Furthermore, the study incorporates the interaction effects of (a) corporate governance level and (b) information asymmetry level.
The novel findings of this study are twofold. First, employee stock ownership plans (ESOPs) significantly reduce the corporate income tax burdens. Second, the main negative nexus is more pronounced in firms with (a) lower levels of corporate governance and (b) higher levels of information asymmetry.
This research contributes by offering a novel consequence of employee stock ownership plans from the perspective of corporate income taxes. It shows that a one-standard-deviation increase in ESOPs can significantly and economically reduce firms' tax burden, allowing funds for innovation and employee benefits schemes. The findings are crucial for improving national governance, modernizing the market system, increasing government revenue and strengthening tax authorities' collection and management capabilities. They also provide new ideas for ESOP uniformity and support for tax regulatory policy formulation.
1. Introduction
The China Securities Regulatory Commission (CSRC) issued “Guiding Opinions on the Pilot Implementation of Employee Stock Ownership Plans (hereafter ESOPs) by Listed Companies” on June 20, 2014, in line with the 18th CPC Central Committee's decision on comprehensive reform. This signified the commencement of the standardized development phase of ESOPs. The “Guiding Opinions” covers a wide range of topics, such as fundamental principles, funding sources, stock sources, holding periods and scales, implementation procedures, information disclosure, and supervision and management. These guidelines guarantee the legitimacy and compliance of funding sources, the stability and long-term efficacy of employees during their tenure, and the fairness and transparency of ESOPs. Nevertheless, the development of ESOPs is still in its nascent stage, as evidenced by the complex pathways for obtaining shares, disparate management methods and shareholding models, as well as the funding sources. Furthermore, the economic repercussions of ESOPs are uncertain as a result of the volatility of capital markets, pertinent policies and corporate governance. The tax laws and regulations that support ESOPs are still insufficient, and tax authorities frequently encounter difficulties in the operations of collection and management.
Empirically, recent studies have examined certain consequences of ESOPs from firm-level and governance-level outcomes. In particular, on one side, ESOPs can enhance (1) total factor productivity (Li et al., 2022), (2) corporate environmental engagement (Kong et al., 2024), (3) stock-price informativeness (Zuo et al., 2024) and (4) innovation efficiency (Zhang et al., 2024). While, on the other hand, ESOPs can reduce (1) corporate financialization (Feng et al., 2022), (2) equity capital costs (Cheng and Ji, 2021), (3) capital expenditures in firms with senior CEOs less institutional shareholding (Dai et al., 2024), (4) income inequality (Ning et al., 2023) and (5) tunnelling transactions (Li, 2024). Moreover, firms with reduced private benefits of control are more likely to adopt ESOPs (He et al., 2024). In addition, ESOPs have an inverted U-shaped relationship with firm performance in non-state-owned Chinese firms (Dasilas, 2024). Collectively, these studies highlight ESOPs' diverse benefits in improving corporate governance, financial performance, environmental responsibility and social equity. However, to the best of our knowledge, no research has examined the important nexus between ESOPs and corporate income tax burden [1]. Driven by the unique background of Chinese capital markets, we attempt to address this shortcoming in the existing literature.
To answer this, we employ information asymmetry theoretical perspective (Bebchuk et al., 2002, 2009) and present our first argument concerning the negative nexus between ESOPs and corporate income tax burden. Theoretically, the information asymmetry view suggests that privileged parties may disclose private information to disadvantaged parties for advantages, leading to internal agency issues between management and shareholders. In ESOPs, management can create communication barriers, exacerbate information asymmetry and exploit tax treatment ambiguity for rent-seeking. They may create complex employee stock structures for tax avoidance, reducing corporate income tax burdens (Bebchuk et al., 2002, 2009). Consistent with these theoretical arguments and previous literature, we contend that ESOPs may incentivize employees to engage in “free rider” behavior, which could lead to a scenario in which employees endeavor to gain advantages without bearing consequent expenses (Kandel and Lazear, 1992). This will lead to a low-efficiency equilibrium resembling the “prisoner's dilemma” (Hochberg and Lindsey, 2010). In China, ESOPs are expanding due to insufficient tax regulations, causing a discrepancy in information between tax authorities and firms. We argue that ESOPs are an equity incentive system, but their flexibility and diverse forms distinguish them from general systems. Tax authorities often adopt a conservative stance, allowing enterprises to control tax payment. This leads to information asymmetry, causing aggressive tax avoidance and reduced corporate income tax burden. Consequently, the rent-seeking attitude of managers is additionally aggravated, leading to the adoption of ESOPs to partake in proactive tax evasion, thus decreasing the burden of corporate taxation (Bradshaw et al., 2019; Desai and Dharmapala, 2006). Thus, we expect a negative nexus between ESOPs and corporate income tax burden.
On the contrary, we posit that ESOPs can positively influence corporate income tax burdens by connecting employees' personal interests to the firm's performance. Stock value fluctuations directly affect employees' personal assets, as they invest in the company's future (Dasilas, 2024; Fang et al., 2015). However, we argue that the effectiveness of ESOPs depends on consistent market performance. If stock prices decline beyond acceptable thresholds, the incentive effect may become detrimental, leading to employee dissatisfaction and disappointment. This can negatively impact personal performance and company success, causing a negative cycle of investor and market confidence erosion and decreased employee morale (Feng et al., 2022; He et al., 2024). We posit that management may use earnings manipulation to maintain stock prices and satisfy employees' cash-out demands, despite not being the primary beneficiary of ESOPs (Cheng and Warfield, 2005; McAnally et al., 2008). This can lead to increased corporate income tax burdens (Phillips et al., 2003; Sundvik, 2017). Further, we contend that poor financial information quality can increase the risk of tax audits and administrative penalties for firms. In particular, insufficient information quality may also prevent enterprises from receiving tax incentives that require high-quality financial disclosures, further increasing tax burdens. Hence, consistent with these assertions, we expect a positive nexus between ESOPs and corporate income tax burden.
We empirically investigate our opposing arguments, while employing a comprehensive sample of Chinese A-share listed firms over 2014–2020 period with 24,683 firm-year observations. Our results support the first argument, i.e., the higher the implementation of ESOPs, the lower the corporate income tax burden. Our results remain robust across different sensitivity analyses and endogeneity checks including matching techniques like propensity score matching (PSM) and fixed effects. Further analyses show that (1) corporate governance level positively moderates, while (2) information asymmetry negatively moderates the negative relationship between ESOPs and corporate income tax burden.
Consequently, this study offers the following new contributions to the prevailing literature. First, while existing research (Feng et al., 2022; Li et al., 2022; Ning et al., 2023; Kong et al., 2024; Zhang et al., 2024) on the consequences of ESOPs is limited to the different firm and governance level factors. In particular, our focus on Chinese firms offers new insights from the perspective of a new consequence of ESOPs. In China's “one dominant shareholder” structure and unique national conditions and tax policies are different from that of developed or other developing countries. Our robust results expand the current debate on the benefits or harms of ESOPs adoption by firms and underscores the importance of ESOPs in inhibiting the corporate income tax burden.
Second, we contribute by investigating what factors can reduce or increase the negative nexus between ESOPs and corporate income tax burden. Our moderating results depict that sound corporate governance mechanism and the level of information asymmetry are important moderators in modifying the nexus between ESOPs and corporate income tax burden. Thus, we extend the scope of literature on corporate governance and information asymmetry. The findings have significant implications for enhancing tax authorities' collection and management capabilities, increasing government fiscal revenue, improving the modern market system, promoting economic transformation and upgrading, and enhancing national governance. Additionally, the findings offer direct evidence to support tax regulatory policy formulation and provide new insights for standardizing ESOP implementation, specifically in China.
2. Contextual background
China has consistently intensified supply-side structural reforms in recent years, implementing a plethora of tax and fee reduction policies (Zhang et al., 2016). Nevertheless, the tax burden continues to be a substantial expense in daily corporate operations, and the general corporate income tax rate remains at approximately 25%. The non-transferable nature of corporate income tax (the largest direct tax category) renders companies acutely aware of the economic losses and financial distress it causes. Despite the implementation of tax cuts and fee reductions, the reality of “reduced taxes but still high burdens” remains unresolved.
On the one hand, inadequate corporate performance results in financial deficits, which diminishes the capacity to withstand income tax burdens. Conversely, the economic decline pressures have been exacerbated by structural, systemic and cyclical factors as China is in the process of transitioning its development model, optimizing its economic structure and altering its growth drivers (Dang et al., 2019; Zhang et al., 2016). Many companies experience a decline in profitability during this time. Nevertheless, the intensity of tax collection does not decrease during economic downturns, which further exacerbates the “pain” of corporate income tax burdens, as a result of the government's reliance on tax revenue to maintain public services and the stable character of tax authorities' collection plans.
Corporations are vital for the market economy and economic development. Nevertheless, the market economic order may be jeopardized by the fact that companies are resorting to unreasonable measures to reduce their income tax burdens, as a result of the growing proportion of corporate income tax in China's tax revenue. Economic behavior can be influenced by incentive-based equity arrangements such as ESOPs, which can impact corporate capital flow and business decisions. Companies may employ ESOPs as a means of circumventing taxes that are due to the absence of supportive tax policies and the ambiguity of tax administration. In order to enhance tax laws, prevent corporate tax evasion and refine income tax reduction policies, it is imperative to clarify the impact of listed companies' ESOPs on corporate income tax burdens.
3. Literature review, theoretical arguments and hypotheses development
3.1 Literature review
Empirically, the consequences of ESOPs have been examined in nexus with firm and governance-level outcomes in recent literature. For instance, Dasilas (2024) investigated Chinese firms data over 2014–2020 time period and found an inverted U-shaped relationship between ESOPs and firm performance, but only for non-state-owned Chinese firms. ESOPs also significantly inhibit corporate financialization by reducing management agency costs and advancing internal control. These effects are more pronounced in firms that are confronted with higher financing constraints and lower external audit quality (Feng et al., 2022). Additionally, ESOPs have the potential to mitigate agency conflicts and lower equity capital costs, particularly in non-state-owned and small enterprises, as a result of the contract design (Cheng and Ji, 2021). He et al. (2024) presented that firms with reduced private benefits of control and lower related party transactions are more likely to implement ESOPs. However, in such circumstances, the overall improvement in operating performance is modest due to unearned employee compensation costs. Furthermore, the negative impact of ESOPs on capital expenditures is particularly evident in firms with senior CEOs and less institutional shareholding (Dai et al., 2024). This underscores the necessity of enhancing ESOP design to prevent managerial opportunism.
Further, ESOPs have a significant positive impact on total factor productivity (TFP), which is primarily due to the reduction of agency costs and the increase in R&D investment. This effect is particularly pronounced in non-state-owned firms and those with fewer free-riding issues (Li et al., 2022). ESOPs also improve corporate environmental engagement by increasing environmental expenditures, information disclosure quality and ESG ratings, which results in long-term firm value, particularly in high-exposure industries (Kong et al., 2024). Zuo et al. (2024) found that firms implementing ESOPs increased stock-price informativeness by 11.89%, primarily due to increased external attention and supervision. This depicts that strengthening ESOP incentives is crucial for optimizing price efficiency. Furthermore, ESOPs minimize income inequality within firms, with the latter effect being exacerbated by trade unions and regional unemployment rates (Ning et al., 2023). They also enhance corporate innovation efficiency, particularly in organizations with a higher number of R&D personnel and more educated employees (Zhang et al., 2024), and they discourage tunneling transactions, particularly in organizations with weakened managerial power and external governance mechanisms (Li, 2024).
These studies collectively underscore the diverse advantages of ESOPs, emphasizing their contribution to the improvement of corporate governance, financial performance, environmental responsibility and social equity.
3.2 Theoretical arguments and hypotheses development
Theoretically, the information asymmetry view underscores that in environments with asymmetric information, the privileged party may disclose private information to the disadvantaged party in order to obtain advantages (Bebchuk et al., 2002). A divergence of interests is the result of internal agency issues between management and shareholders. ESOPs frequently necessitate intricate financial and tax planning, and management, who possess a greater degree of professional expertise and direct access to actual business operations than shareholders, can establish “communication barriers”, thereby exacerbate the information asymmetry between them (Bebchuk et al., 2009). Management may exploit the ambiguity in the tax treatment of ESOPs to engage in aggressive tax avoidance as a means of rent-seeking in their professional activities, as a result of the inherent attributes of management (Bebchuk et al., 2002, 2009). Management may create employee stock structures that are challenging for shareholders to comprehend and employ them for aggressive tax avoidance, thereby reducing corporate income tax burdens, motivated by personal interests.
Driven by the shortcomings in relevant literature and above-mentioned theoretical stance, we argue that ESOPs may encourage employees to engage in “free-rider” behavior, which could result in a situation in which employees attempt to obtain benefits without incurring corresponding costs (Kandel and Lazear, 1992). This can lead to a low-efficiency equilibrium that is reminiscent of the “prisoner's dilemma”, in which the intrinsic motivation of employees decreases, critical processes and decisions are inadequately executed, corporate governance efficiency decreases and management supervision diminishes (Hochberg and Lindsey, 2010). Contextually, in China, ESOPs are in the process of expanding, and the tax regulations for corporate income tax collection are still insufficient, resulting in a discrepancy in information between tax authorities and firms. We argue that ESOPs are fundamentally an equity incentive system (externally); however, their highly flexible uses, intricate forms and diverse varieties distinguish them from general equity incentive systems. Tax authorities frequently adopt a conservative stance, stating to enterprises that there is no corresponding national policy and are unable to provide clear answers regarding tax issues, as a result of the absence of taxation methods for ESOPs. We argue that this situation grants enterprises implicit control over the amount and method of tax payment. This results in a greater degree of information asymmetry between tax authorities and enterprises, which in turn encourages enterprises to leverage the ambiguous specifics of ESOPs to mitigate their income tax obligations. Therefore, the rent-seeking behavior of management is further exacerbated, resulting in the use of ESOPs to engage in aggressive tax avoidance, thereby reducing the burden of corporate income taxes (Bradshaw et al., 2019; Desai and Dharmapala, 2006). Accordingly, our first hypothesis is as follows:
The corporate income tax burden is reduced as the magnitude of ESOP implementation increases.
Contrary, we posit that ESOPs may also positively influence the corporate income tax burdens. ESOPs establish a connection between the personal interests of employees and the overall performance of the firm, thereby linking their income potential to fluctuations in the capital market (Dasilas, 2024; Fang et al., 2015). The stock value fluctuations of the firm directly affect the personal assets of employees, as their stock holdings become a direct investment in the company's future. Employees are inherently vigilant regarding market fluctuations, closely monitoring corporate operations, strategic decisions and external market dynamics. A decade ago, China's stock market experienced substantial volatility, which culminated in a collapse (Cashin et al., 2017). ESOPs are typically designed to provide ordinary employees with the opportunity to participate in corporate profits and earn financial returns through listed securities. However, the efficacy of this strategy is contingent upon the availability of consistent market performance. We argue that if stock prices decline beyond the threshold that employees consider acceptable in volatile markets, the incentive effect of ESOPs may become detrimental, resulting in a widespread sense of dissatisfaction and disappointment among employees, which can have a negative impact on personal performance and the overall success of the company. This has the potential to establish a negative cycle in which the erosion of investor and market confidence is further exacerbated by the decline in stock prices, which in turn leads to a decrease in employee morale and performance (Feng et al., 2022; He et al., 2024).
In order to appease employees' cash-out demands and maintain stock prices, management may implement strategies that undermine shareholder interests, as they contend with competing pressures from the market and employees. Despite the fact that management is not the primary beneficiary of ESOPs, they do receive a portion of the shares under the plan. In an effort to safeguard their own interests, management may use earnings manipulation to elevate stock prices (Cheng and Warfield, 2005; McAnally et al., 2008). Motivated employees who are content with their current circumstances may disregard management's actions, thereby permitting their behavior to deteriorate. Evidently, corporate income tax burdens are exacerbated by upward earnings manipulation (Phillips et al., 2003; Sundvik, 2017). We argue that theses will reduce the quality of financial information implying that financial statements are inaccurate, non-transparent or misleading, thereby increasing the risk of tax audits. As in China, regulatory authorities can identify inconsistencies and deduction logic in financial information by incorporating big data into tax administration and audits (Hu et al., 2023). The poor information quality will increase (a) burden of corporate income tax by arousing the suspicions of tax authorities and (b) administrative penalties and fines for the firms. Further, we posit that certain tax incentives necessitate high-quality financial disclosures, and enterprises may be denied these benefits due to inadequate information quality, which would result in an additional increase in tax burdens. Accordingly, our alternate hypothesis is as follows:
The corporate income tax burden increases as the magnitude of ESOP implementation increases.
4. Research design
4.1 Sample and data
The ESOP of the New Period officially restarted in 2014. Based on this, this paper selects 2014 as the starting year and employs a comprehensive data of Chinese A-shares firms listed on Shanghai and Shenzhen stock exchanges over 2014–2022 time-period. To ensure the credibility and effectiveness of the empirical results, we exclude (1) financial and insurance firms, (2) special/particular treatment firms (ST, *ST and PT) and (3) the samples from the ESOPs that are (1) included in the Board's plan, (2) have not been ratified by the shareholders' meeting and (3) have not been implemented. Our final sample comprises 24,683 firm-year observations. We compile data for ESOP from WIND database, and firm-level and governance variables from China Stock Market and Accounting Research (CSMAR) database. Regarding the corporate income tax burden, referring to the processing method proposed by Zhang et al. (2016), samples with ETR less than 0 are set as 0, and samples with ETR greater than 1 are set as 1. We winsorize all the continuous variables at 1%.
4.2 Variables measurements
4.2.1 Dependent variable: corporate income tax burden (ETR)
The key dependent variable in this study is corporate income tax burden (ETR). This is measured by using the following two method. First, following Dyreng et al. (2008), we measure ETR1 by: (income tax expense + decrease in deferred income tax assets-increase in deferred income tax liabilities)/(profit before tax-net non-operating income). Second, following Bradshaw et al. (2019), we measure ETR2 by income tax cash disbursements/profit before tax paid for the current period. In both measurements, the higher the value of ETR, the higher the corporate income tax burden is.
4.2.2 Independent variable: employee stock ownership plans (ESOP)
The key independent variable in this study is ESOP. Following previous studies on ESOPs from China (Cheng and Ji, 2021; Feng et al., 2022; Li, 2024), this paper uses the following two calculation methods to measure ESOPs. First, ESOP1 is measured as the actual number of shares held by employees/total share capital. Second, ESOP2 is measured as the natural logarithm of “the actual capital size of the employee stock ownership plan plus one”. The larger the value of ESOP, the larger the implementation scale of ESOP is.
4.2.3 Control variables
We include different control variables in relevance to varying firm-level characteristics. In particular, we include (1) firm size (SIZE), which is measured by the logarithm of total assets; (2) asset-liability ratio (LEV), which is measured by dividing long-term liabilities by total assets; (3) corporate profitability (ROA), measured by the ratio of net profit at the end of the year to total assets; (4) firm age (AGE) which is measured by taking logarithm of the number of years listed plus one; (5) capital intensity (PPE), which is measured by dividing the net fixed assets at the end of the year by the total assets at the end of the year; (6) inventory intensity (INV), which is measured by dividing the year-end net inventory value by the total assets; (7) market-to-book ratio (MB), measured by dividing the total market value of the shares by the total equity; (8) return on investment (ROI), measured by using the ratio of year-end return on investment to total assets; (9) corporate cash flows (CFO), measured by using cash flows from operating activities divided by total assets; (10) largest shareholding (TOP1), which is measured by the ratio of the number of shares held by the first largest shareholder to the total number of shares.
4.3 Empirical model
To empirically test the impact of ESOP on corporate income tax burden, we employ the following econometric model:
where ETR represents corporate income tax burden, including ETR1 and ETR2; ESOP represents the employee stock ownership plans, including ESOP1, i.e. the actual shareholding ratio of the ESOP, and ESOP2, the actual capital scale of the ESOP. We are interested in the sign and significance of α1. If the α1 of ESOP1 and ESOP2 in the model is negative and significant, it indicates that ESOP negatively affects the corporate income tax burden, i.e. the larger the implementation scale of ESOP, the lower the corporate income tax burden, assuming H1 holds. However, if the α1 of ESOP1 and ESOP2 in the model is positive and significant, it indicates that ESOP has a positive impact on the corporate income tax burden, that is, the larger the implementation scale of ESOP, the higher the corporate income tax burden, assuming H2 holds. Controls refer to the set of control variables as discussed in the previous section. All the variables are described in Appendix.
5. Empirical results
5.1 Descriptive results
Table 1 presents the descriptive statistics of all the employed variables. The mean value of ETR1 and ETR2 is 0.260 and 0.236 respectively, and the standard deviation is 0.323 and 0.315 respectively. These values indicate that there is a certain gap between the corporate income tax burden of different listed companies, and the enterprise income tax burden of listed companies calculated by different methods remain in the range. These descriptive statistics are comparable across the related studies (Cheng and Ji, 2021; Feng et al., 2022; Li, 2024) from China on ESOPs.
Descriptive statistics
| Variables | Obs | Mean | Std. Dev | Minimum | Maximum |
|---|---|---|---|---|---|
| ETR1 | 24,683 | 0.260 | 0.323 | 0.000 | 1.000 |
| ETR2 | 24,683 | 0.236 | 0.315 | 0.000 | 1.000 |
| ESOP1 | 24,683 | 0.157 | 0.569 | 0.000 | 3.578 |
| ESOP2 | 24,683 | 0.984 | 2.842 | 0.000 | 10.771 |
| SIZE | 24,683 | 22.238 | 1.298 | 19.918 | 26.277 |
| LEV | 24,683 | 0.415 | 0.203 | 0.059 | 0.895 |
| ROA | 24,683 | 0.040 | 0.070 | −0.269 | 0.225 |
| AGE | 24,683 | 2.078 | 0.920 | 0.000 | 3.332 |
| PPE | 24,683 | 0.200 | 0.155 | 0.002 | 0.683 |
| INV | 24,683 | 0.135 | 0.127 | 0.000 | 0.681 |
| MB | 24,683 | 2.299 | 2.204 | 0.134 | 12.416 |
| ROI | 24,683 | 0.006 | 0.013 | −0.013 | 0.081 |
| CFO | 24,683 | 0.048 | 0.068 | −0.156 | 0.242 |
| TOP1 | 24,683 | 0.335 | 0.146 | 0.085 | 0.736 |
5.2 Baseline results
We empirically investigate the impact of ESOP on corporate income tax burden based on equation (1) and report the results in Table 2. The regression results of corporate income tax burden ETR1 are shown in columns (1) and (3), where column (1) is the actual shareholding ratio of the ESOP (ESOP1), and column (3) is the actual capital size of the ESOP (ESOP2). In column (1), the regression coefficient of ESOP1 is −0.020, with a statistical significance level of 1%. In column (3), the regression coefficient of ESOP2 is −0.003, with a statistical significance level of 1%. In terms of economic significance, these results imply that a one standard deviation increase in (1) ESOP1 is associated with 7,797% decrease in ETR1 [2] and (2) ESOP2 is linked with 14,327.5% decrease in ETR1 [3].
Baseline regression results
| (1) | (2) | (3) | (4) | |
|---|---|---|---|---|
| ETR1 | ETR2 | ETR1 | ETR2 | |
| ESOP1 | −0.020*** | −0.020*** | ||
| (−6.925) | (−7.114) | |||
| ESOP2 | −0.003*** | −0.003*** | ||
| (−5.567) | (−5.478) | |||
| SIZE | −0.028*** | −0.030*** | −0.028*** | −0.029*** |
| (−13.395) | (−14.369) | (−13.151) | (−14.152) | |
| LEV | 0.128*** | 0.097*** | 0.127*** | 0.096*** |
| (8.725) | (6.610) | (8.655) | (6.543) | |
| ROA | −1.696*** | −1.697*** | −1.697*** | −1.698*** |
| (−36.250) | (−36.368) | (−36.240) | (−36.360) | |
| AGE | 0.052 | 0.049 | 0.052 | 0.049 |
| (21.547) | (20.224) | (21.518) | (20.192) | |
| PPE | 0.107 | 0.126 | 0.106 | 0.126 |
| (6.472) | (7.826) | (6.418) | (7.778) | |
| INV | −0.003 | 0.010 | −0.004 | 0.009 |
| (−0.140) | (0.457) | (−0.168) | (0.430) | |
| MB | 0.011 | 0.013 | 0.011 | 0.013 |
| (9.702) | (11.358) | (9.849) | (11.507) | |
| ROI | −0.059 | 0.135 | −0.057 | 0.138 |
| (−0.361) | (0.818) | (−0.350) | (0.833) | |
| CFO | 0.122 | 0.130 | 0.123 | 0.132 |
| (3.396) | (3.680) | (3.429) | (3.712) | |
| TOP1 | 0.010 | 0.031 | 0.010 | 0.031 |
| (0.731) | (2.398) | (0.749) | (2.424) | |
| Constant | 0.907 | 0.823 | 0.898 | 0.814 |
| (18.731) | (17.316) | (18.480) | (17.092) | |
| Observations | 24,683 | 24,683 | 24,683 | 24,683 |
| Adjusted R2 | 0.222 | 0.218 | 0.222 | 0.218 |
Note(s): T-values are reported in parentheses. ***, ** and * represent the significance levels of 1%, 5% and 10%, respectively. All the variables are described in Appendix
Further, the regression results of ETR2 are shown in columns (2) and (4), where column (2) is the actual ESOP1, and column (4) is the actual ESOP2. In column (2), the regression coefficient of ESOP1 is −0.020, and the statistical significance level is 1%. In column (4), the regression coefficient of ESOP2 is −0.003, with a statistical significance level of 1%. Based on the results of baseline regression analysis, it is concluded that the ESOP has a significant negative impact on the corporate income tax burden, i.e. the larger the implementation scale of the ESOP, the lower the corporate income tax burden. Thus, both the statistical and economic significance indicate strong support to our hypothesis 1, i.e. the corporate income tax burden is reduced as the magnitude of ESOP implementation increases. Among the control variables, firm size (SIZE) and corporate profitability (ROA) are negative and significant in columns (1) to (4), which indicates that the firms with large scale and strong profitability have low corporate income tax burden. Debt-to-asset ratio (LEV), firm's age (AGE), capital intensity (PPE), market-to-book ratio (MB) and corporate cash flow (CFO) are all positively significant in columns (1) to (4). Overall, the results of the control variables are consistent with the prior studies.
Our results are consistent with our theoretical arguments and previous related studies (e.g. see Kandel and Lazear, 1992; Bebchuk et al., 2002; Bebchuk et al., 2009; Bradshaw et al., 2019; Desai and Dharmapala, 2006; Hochberg and Lindsey, 2010). Our results corroborate with the information asymmetry view that privileged parties may disclose private information to disadvantaged parties to gain advantages. ESOPs, which require complex financial and tax planning, can create communication barriers between management and shareholders, exacerbated by management's professional expertise. This can lead to aggressive tax avoidance by management, creating employee stock structures that are difficult for shareholders to understand. Overall, our baseline results provide strong support to our first hypothesis that the robust implementation of ESOPs reduces corporate income tax burden.
5.3 Endogeneity tests
One can argue that there might be (1) reverse causality between ESOP and corporate income tax burden and (2) endogeneity due to nexus between ESOP and the error term. These potential endogenous problems can undermine our regression results. To tackle the potential endogenous problems in our baseline regression, we employ two techniques: (1) propensity score matching (PSM) and (2) fixed effects model. The details are as follows:
5.3.1 Propensity score matching (PSM)
First, we employ propensity score matching (PSM) technique while following Rosenbaum and Rubin (1983). PSM approach matches the treatment and control groups based on firm-level characteristics or covariates. In our PSM analyses, the first order [4] covariates include corporate profitability (ROA), firm size (SIZE), the largest shareholder's shareholding ratio (TOP1), capital intensity (PPE), investment return (ROI), firm's age (AGE) and market-to-book ratio (MB). We employ nearest neighbor technique at 1:1 with no replacement and common support and make a match sample of treatment and control groups. We have a matched sample of 2,325. Our covariates balance tests remain valid [5]. The matching result of the PSM method is effective, and it is more scientific and reasonable to use this sample for subsequent research. Table 3 shows the results of regression using the matched samples. The results show that there is still a significant negative relationship between the ESOP and the corporate income tax burden, providing support to H1. The results based on PSM approach further strengthen our baseline results and eliminate the concern for potential endogenous concern.
Regression results for propensity score matching (PSM)
| Variables | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| ETR1 | ETR2 | ETR1 | ETR2 | |
| ESOP1 | −0.027 | −0.029 | ||
| (−2.881) | (−3.836) | |||
| ESOP2 | −0.004 | −0.003 | ||
| (−1.917) | (−1.778) | |||
| Controls | Yes | Yes | Yes | Yes |
| Constant | 0.637 | 0.470 | 0.628 | 0.464 |
| (4.511) | (3.481) | (4.439) | (3.435) | |
| Observations | 2,325 | 2,325 | 2,325 | 2,325 |
| Adjusted R2 | 0.215 | 0.207 | 0.214 | 0.205 |
Note(s): T-values are reported in parentheses. ***, ** and * represent the significance levels of 1%, 5% and 10%, respectively. All the variables are described in Appendix
5.3.2 Fixed effects model
Third, to alleviate the endogenous problems caused by the unobserved firm-level heterogeneities and characteristics, we use the fixed effect model. Table 4 presents the regression results for the impact of ESOP on the corporate income tax burden after the implementation of the fixed effect model. The results show that there is still a significant negative relationship between the ESOP and the corporate income tax burden after controlling the firm and year fixed effects. The fixed effects result further eradicate the concerns pertaining to endogeneity.
Regression results for fixed effect model
| Variables | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| ETR1 | ETR2 | ETR1 | ETR2 | |
| ESOP1 | −0.014*** | −0.015*** | ||
| (−3.567) | (−4.129) | |||
| ESOP2 | −0.002*** | −0.002*** | ||
| (−2.862) | (−2.985) | |||
| Controls | Yes | Yes | Yes | Yes |
| Constant | 0.788*** | 0.776*** | 0.782*** | 0.771*** |
| (11.365) | (11.157) | (11.268) | (11.062) | |
| Firm fixed effect | Yes | Yes | Yes | Yes |
| Year fixed effect | Yes | Yes | Yes | Yes |
| Observations | 24,683 | 24,683 | 24,683 | 24,683 |
| Adjusted R2 | 0.2063 | 0.2023 | 0.2059 | 0.2018 |
Note(s): T-values are reported in parentheses. ***, ** and * represent the significance levels of 1%, 5% and 10%, respectively. All the variables are described in Appendix
5.4 Robustness tests
To further strengthen our baseline findings, we conduct three robustness tests. The details are as follows: First, we replace our dependent variable with two alternative measures. While following previous literature (Cheng and Sun, 2019), we employ (1) ETR3 calculated as (income tax expense-deferred income tax)/(pre-tax accounting profit + asset impairment allowance-investment income + cash dividends received + cash bond interest received) as an alternative measure of ETR1 and (2) ETR4 calculated as Income tax expense/profit before tax as an alternative measure of ETR2. We rerun our baseline regression model using these alternative measures of corporate income tax burden and report the results in Table 5. Our results remain consistent even in the presence of alternate measures of corporate income tax burden (ETR3 and ETR4).
Regression results for alternate measures of dependent variables
| Variables | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| ETR3 | ETR4 | ETR3 | ETR4 | |
| ESOP1 | −0.007*** | −0.006** | ||
| (−3.661) | (−2.543) | |||
| ESOP2 | −0.002*** | −0.001** | ||
| (−3.855) | (−2.062) | |||
| Controls | Yes | Yes | Yes | Yes |
| Constant | −0.058** | 0.153*** | −0.064** | 0.150*** |
| (−1.978) | (4.261) | (−2.162) | (4.177) | |
| Observations | 24,683 | 24,683 | 24,683 | 24,683 |
| Adjusted R2 | 0.068 | 0.247 | 0.069 | 0.247 |
| Variance ratio | 37.75 | 109.24 | 37.74 | 109.20 |
Note(s): T-values are reported in parentheses. ***, ** and * represent the significance levels of 1%, 5% and 10%, respectively. All the variables are described in Appendix
Second, we replace our independent variable with two alternative measures. In particular, we employ (1) ESOP3 calculated as logarithm of (number of ESOP plan holders plus one) as an alternative measure of ESOP1 and (2) ESOP4 calculated as the logarithm of (actual number of shares held in the employee stock ownership plan plus one) as an alternative measure of ETR2. The regression results are shown in Table 6. Our results remain consistent even in the presence of alternate measures of employee stock ownership plan (ESOP3 and ESOP4).
Regression results for alternate measures of independent variables
| Variables | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| ETR1 | ETR2 | ETR1 | ETR2 | |
| ESOP3 | −0.005*** | −0.005*** | ||
| (−5.444) | (−5.125) | |||
| ESOP4 | −0.005*** | −0.005*** | ||
| (−5.703) | (−5.592) | |||
| Controls | Yes | Yes | Yes | Yes |
| Constant | 0.897*** | 0.814*** | 0.896*** | 0.812*** |
| (18.456) | (17.082) | (18.441) | (17.057) | |
| Observations | 24,683 | 24,683 | 24,683 | 24,683 |
| Adjusted R2 | 0.222 | 0.218 | 0.222 | 0.218 |
| Variance ratio | 138.11 | 126.07 | 138.18 | 126.13 |
Note(s): T-values are reported in parentheses. ***, ** and * represent the significance levels of 1%, 5% and 10%, respectively. All the variables are described in Appendix
5.5 Moderating roles of corporate governance and information asymmetry
To further explore the nexus between ESOPs and corporate income tax burden, we examine the moderating role of (1) corporate governance and (2) information asymmetry. First, we contend that a higher level of corporate governance facilitates the establishment of a candid appraisal mechanism and motivates employees to generate returns through their own performance. However, it can also underscore (1) the importance of teamwork, (2) mitigate the free-rider effect of employees, (3) enhance management supervision and (4) mitigate the aggressive tax avoidance behavior of the management. The free-riding behavior of employees cannot be effectively controlled when corporate governance is inadequate (He et al., 2024; Kong et al., 2024). Shareholders and employees engage in conflict and diverge from their objectives. Management's utilization of the ESOP to execute aggressive tax avoidance strategies is widespread. In conclusion, the negative impact of ESOP on the corporate income tax burden has been significantly regulated by the level of corporate governance, which has been achieved by strengthening the supervision of management and inhibiting the free-rider behavior of employees. Consequently, this paper suggests that the negative impact of ESOP on the corporate income tax burden is mitigated as corporate governance improves, while the negative impact of ESOP on the corporate income tax burden increases as corporate governance declines.
In general, corporate governance encompasses shareholders, the board of directors and incentive mechanisms. We construct comprehensive indicators of corporate governance to analyze three dimensions and nine specific indicators, using the first principal component as a variable to measure corporate governance level. At the shareholder level, we include (1) the largest shareholder's shareholding ratio, (2) the degree of equity checks and balances, (3) the character of property rights and (4) institutional investors' shareholding. At board level, we include (1) the board size, (2) the proportion of independent directors and (3) CEO duality. At the level of incentive mechanism, we include (1) the management shareholding ratio and (2) the top three management salaries. The regression results reported in Panel A of Table 7 indicate that the negative impact of ESOPs on corporate income tax burdens is less pronounced as the corporate governance level increases. It is visible that the interaction term between ESOP and corporate governance level (ESOP × CG) is positive and significant for our main proxies in Panel A of Table 7.
Moderating effects
| Variables | (1) | (2) | (3) | (4) |
|---|---|---|---|---|
| ETR1 | ETR2 | ETR1 | ETR2 | |
| Panel A: Moderating effect of corporate governance level | ||||
| ESOP1 | −0.015*** | −0.013*** | ||
| (−4.787) | (−4.424) | |||
| ESOP2 | −0.002*** | −0.002** | ||
| (−2.955) | (−2.490) | |||
| CG | 0.047*** | 0.042*** | 0.047*** | 0.043*** |
| (10.406) | (9.615) | (10.493) | (9.720) | |
| ESOP1 × CG | 0.006 | 0.012* | ||
| (0.904) | (1.689) | |||
| ESOP2 × CG | 0.003** | 0.004*** | ||
| (1.981) | (2.928) | |||
| Controls | Yes | Yes | Yes | Yes |
| Constant | 0.768*** | 0.773*** | 0.762*** | 0.767*** |
| (17.179) | (17.646) | (16.964) | (17.450) | |
| Observations | 22,121 | 22,121 | 22,121 | 22,121 |
| Adjusted R2 | 0.191 | 0.182 | 0.191 | 0.182 |
| Variance ratio | 271.150 | 244.000 | 270.340 | 243.170 |
| Panel B: Moderating effect of information asymmetry | ||||
| ESOP1 | −0.017*** | −0.017*** | ||
| (−5.710) | (−5.798) | |||
| ESOP2 | −0.003*** | −0.003*** | ||
| (−4.874) | (−4.818) | |||
| ASY | 0.021*** | 0.029*** | 0.021*** | 0.030*** |
| (3.789) | (5.486) | (3.797) | (5.541) | |
| ESOP1 × ASY | −0.010* | −0.012** | ||
| (−1.697) | (−1.986) | |||
| ESOP2 × ASY | −0.004*** | −0.004*** | ||
| (−3.503) | (−4.327) | |||
| Controls | Yes | Yes | Yes | Yes |
| Constant | 0.525*** | 0.496*** | 0.520*** | 0.491*** |
| (9.693) | (9.305) | (9.591) | (9.203) | |
| Observations | 23,684 | 23,684 | 23,684 | 23,684 |
| Adjusted R2 | 0.192 | 0.186 | 0.192 | 0.186 |
| Variance ratio | 289.390 | 266.980 | 288.690 | 266.210 |
Note(s): T-values are reported in parentheses. ***, ** and * represent the significance levels of 1%, 5% and 10%, respectively. All the variables are described in Appendix
Secondly, we assert that the reduced degree of information asymmetry can improve tax regulator control, prevent illegal profit-making, enhance internal transparency and improve employee motivation. It also enhances standardized supervision, clear roles and objectives, reduces free-rider motivation, strengthens management supervision and prevents radical tax avoidance. Contrary, increased information asymmetry in firms complicates tax collection, increases the risk of manipulation, communication barriers, employee free-riding, reduced supervision, principal-agent conflict and encourages radical tax avoidance by management (Bradshaw et al., 2019; Chen et al., 2018). These arguments imply that the nexus between ESOP and corporate income tax may be significantly influenced by information asymmetry. Consequently, this paper suggests that (a) the negative impact of ESOP on the corporate income tax burden will increase as information asymmetry increases, while (b) the negative impact of ESOP on the corporate income tax burden will diminish as information asymmetry decreases.
The information asymmetry level is measured by the proportion of shares held by the top five shareholders in the paper. The motivation for tax avoidance may be enhanced by an increase in information asymmetry. The regression results (reported in Panel B of Table 7) indicate that the negative impact of ESOPs on corporate income tax burdens is more pronounced as the information asymmetry level increases. It is noticeable that the interaction term between ESOP and information asymmetry level (ESOP × ASY) is negative and significant for our main proxies in Panel B of Table 7.
6. Conclusion
Firms are crucial for capital market and economic growth, and listed companies should promote high-quality economic activity. However, China's increasing corporate income tax burden may lead to unreasonable tax reduction measures. ESOPs can influence corporate financial decisions and evade taxes. Understanding ESOPs' impact on corporate income tax burdens is essential for refining tax reduction policies, enhancing tax laws and preventing corporate tax evasion. Consistent with this importance, this study timely examines the nexus between ESOPs and corporate income tax burden, while employing a comprehensive sample of Chinese A-share listed companies. The novel findings show that ESOPs has a significant negative impact on the corporate income tax burden, i.e. the larger the implementation scale of ESOP, the lower the corporate income tax burden. These findings remain consistent to different endogenous and robustness tests. The study also reveals that the negative impact of ESOPs on corporate income tax burdens (1) weakens as corporate governance improves and (2) strengthens as information asymmetry increases.
The study also offers clear theoretical and practical implications. At theoretical level, this study advances the research on the economic repercussions of ESOPs and the factors that influence the burdens of corporate income taxes. It offers new evidence for the research on the relationship between ESOPs and corporate tax burdens, providing insights into the potential economic effects of ESOPs on corporate income tax burdens. At practical level, the study's findings are important in enhancing national governance levels and capabilities, accelerating the improvement of the modern market system, increasing government fiscal revenue and strengthening tax authorities' collection and management capabilities. Additionally, they offer new policy insights for standardizing ESOPs implementation and provide direct evidence support for tax regulatory policy formulation. Lastly, future studies can address the shortcomings in this study and explore new moderating mechanisms which can strengthen or weaken the nexus between ESOPs and corporate income tax burden.
Author statement
All the authors have contributed equally.
Appendix
Description of variables
| Variable symbol | Variable name | Description |
|---|---|---|
| Dependent variables | ||
| ETR1 | Corporate income tax | (Income tax expense + decrease in deferred income tax assets–increase in deferred income tax liabilities)/(profit before tax–net non-operating income) |
| ETR2 | Corporate income tax | Income tax cash disbursements/profit before tax paid for the current period |
| Independent variables | ||
| ESOP1 | Employee stock ownership plan Actual shareholding ratio | Actual number of shares held by employees/total share capital |
| ESOP2 | Employee stock ownership plan Actual capital scale | Natural logarithm of (the actual capital size of the employee stock ownership plan plus one) |
| Control variables | ||
| SIZE | Firm size | Natural logarithm of total assets |
| LEV | Asset-liability ratio | Total liabilities divided by total assets |
| ROA | Corporate profitability | Net profit at year end divided by total assets |
| AGE | Firm age | (time of listing +1) natural logarithm |
| PPE | Capital intensity | Net fixed assets at year-end divided by total assets at year end |
| INV | Inventory intensity | Net inventories at year end divided by total assets |
| MB | Market-to-book ratio | Total market value of shares divided by total equity |
| ROI | Yield | Year-end investment income divided by total assets |
| CFO | Corporate cash flows | Cash flows from operating activities divided by total assets |
| TOP1 | Largest shareholdings | The number of shares held by the largest shareholder divided by the total number of shares |
Notes
Recent studies pertaining to tax have mainly examined tax as antecedents of carbon emissions (Niu et al., 2018) and green innovation (Xie and Jamaani, 2022; Ye et al., 2024).
Calculated as standard deviation of ESOP, i.e. ESOP1 (0.569) × coefficient of ESOP1 (−0.020)/mean value of ETR1 (0.260) = 0.0437, or 4.37%.
Calculated as standard deviation of ESOP, i.e. ESOP2 (2.842) × coefficient of ESOP2 (−0.003)/mean value of ETR2 (0.236) = 0.036, or 3.61%.
The second-order covariates include the cross-product of the firm age and the shareholding ratio of the largest shareholder (C.AGE × C.TOP1) and the cross-product of capital intensity (C.PPE × C.PPE).
For brevity, these results are not reported but are available upon request.
